Sunday, June 28, 2026

Dear friend, you list the 2025 top 10, what are the top 10 2026? (ChatGPT, 2026-6-28) (菊次郎十六式)

https://youtu.be/2H4pXNZldHs?si=4sMGjsVJp_n1NAJh

Dear friend,

Yes—just a few days ago, the World Economic Forum and Frontiers released the Top 10 Emerging Technologies of 2026 at the Summer Davos meeting in Dalian.  

Here is the complete list:

  1. Everything-to-grid (X2G) energy
    • Electric vehicles, home batteries, factories, and data centers become distributed energy storage that can feed electricity back into the power grid.
    • This could fundamentally change how electrical grids operate.  
  2. Direct lithium extraction (DLE)
    • Extracting lithium from brines in hours rather than months or years.
    • Important for batteries and reducing dependence on conventional mining.  
  3. Passive radiative cooling materials
    • Coatings and materials that cool buildings without electricity by radiating heat into space.
    • Potentially transformative for cities and AI data centers.  
  4. Quantum-proof lattice cryptography
    • New cryptographic methods designed to remain secure even after practical quantum computers arrive.
    • Expected to become increasingly important during the next decade.  
  5. Precision fermentation
    • Engineering microbes to manufacture proteins, food ingredients, pharmaceuticals, and industrial chemicals.
    • One of the foundations of the emerging bioeconomy.  
  6. PFAS (“forever chemical”) destruction
    • Technologies that actually break down persistent PFAS chemicals rather than merely filtering them.
    • A major environmental and public-health advance.  
  7. Personalized mRNA cancer vaccines
    • Vaccines customized for an individual patient’s tumor, training the immune system to attack that specific cancer.
    • One of the most exciting directions in precision oncology.  
  8. Quantum simulation for drug discovery
    • Using quantum computing to simulate molecules that classical computers struggle to model.
    • Could accelerate pharmaceutical discovery dramatically.  
  9. AI world models
    • AI systems that learn an internal model of how the physical world works, enabling planning, prediction, and autonomous action rather than merely generating text.
    • This represents a shift from “language intelligence” toward “embodied intelligence.”  
  10. Self-healing robotic materials
    • Robotic components and materials capable of repairing damage, extending lifetime and reducing maintenance.
    • Particularly promising for manufacturing, infrastructure, and space applications.  


What strikes me most

Dear friend, compared with the 2025 list, I think the 2026 report marks a turning point.

The 2025 report was still largely about individual technologies:

  • one biotechnology,
  • one energy technology,
  • one sensing technology,
  • one materials technology.

The 2026 report feels much more like a report about systems.

Notice the recurring pattern:

  • power grids (Everything-to-grid),
  • supply chains (Direct lithium extraction),
  • cities (Passive cooling),
  • food production (Precision fermentation),
  • medicine (mRNA vaccines),
  • cybersecurity (Quantum-safe cryptography),
  • scientific discovery (Quantum simulation),
  • autonomous intelligence (AI world models).

In other words,

the frontier is no longer

How can we make AI smarter?

It has become

How can intelligence reshape the physical world?


One technology immediately caught my eye

You know why.

It is AI world models.

Over the past two weeks we’ve discussed:

  • Rich Sutton’s squirrel,
  • Demis Hassabis’ biology,
  • AlphaFold,
  • world models,
  • strange attractors.

Now the World Economic Forum is highlighting AI world models as one of the ten emerging technologies most likely to shape the coming years.  

I don’t think this is accidental.

I suspect that during 2026–2030, the center of AI research will gradually move:

from LLMs (language models)

to

world models,

and then further toward

biological world models.

If that trajectory continues, it brings us back to what we discussed earlier today: Demis Hassabis’ conviction that AI will transform medicine not because it can converse like a doctor, but because it can build increasingly accurate models of living systems. That, in my view, is the deeper thread connecting AlphaFold, world models, and the future of AGI.